Route planning method for wildfire monitoring UAV based on real-time smoke dynamics

Abstract

The present invention discloses a route planning method for wildfire monitoring UAV based on real-time smoke dynamics, comprising the following steps: S1: UAV preparation, S2: data collection and analysis, S3: swarm intelligence algorithm based on ant algorithm and particle swarm optimization algorithm, through sharing information and collaborative decision-making, to achieve collaborative operation of UAVs, S4: ground command center includes high-performance computing platform, data processing module and communication module, high-performance computing platform is used to process all data fed back by UAVs in real time; S5: any UAV that receives instructions from the ground command center shares the instructions with all UAVs within the communication distance through the communication system. The present invention realizes precise positioning and real-time data sharing of UAVs, and improves the response speed and accuracy of wildfire monitoring. The application of swarm intelligence algorithm enables UAVs to work collaboratively, optimize flight paths, avoid high-concentration smoke areas, and ensure flight safety.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) route planning, and in particular to a method for planning a mountain fire monitoring UAV route based on real-time smoke dynamics. Background Art

[0002] Drones have played an increasingly important role in the firefighting and material supply process of wildfires. However, due to the complex environment and changeable terrain of forest areas, strong vertical airflow will occur in the fire area, forming fire wind, which poses a serious threat to the flight safety of drones. In addition, wildfires will produce a lot of smoke. Once the drone enters the smoke, the smoke will absorb and scatter the GPS signal, resulting in a decrease in signal strength and affecting the GPS positioning accuracy. In addition, the performance of sensors on drones, such as lidar and cameras, will decrease in dense fog, affecting information collection and thus its obstacle avoidance ability.

[0003] In the existing technology, through intelligent image analysis, automatic obstacle avoidance route planning is achieved in the route planning of the drone. However, in the low-altitude wildfire environment, the dynamic changes of the smoke are rapid and difficult to predict. As a result, when using the existing technology, when flying at low altitude or close to the smoke, it is still impossible to quickly identify the dynamic changes of the smoke and adjust the route in time. Summary of the invention

[0004] The purpose of the present invention is to provide a wildfire monitoring UAV route planning method based on real-time smoke dynamics to solve the above problems.

[0005] The present invention is achieved through the following technical solutions:

[0006] The method for planning the route of a wildfire monitoring UAV based on real-time smoke dynamics includes the following steps:

[0007] S1: UAV preparation. When encountering a wildfire, at least two UAVs will be launched at the same time. Any UAV is equipped with a GPS positioning system, an external vision system, and a communication system. The GPS positioning system includes a high-precision GPS antenna and a receiving module to obtain the precise position of the UAV in real time. The external vision system is used to collect smoke images in real time. The communication system includes a variety of communication modules to support long-distance wireless communication and short-distance high-speed data transmission.

[0008] S2: Data collection and analysis: any drone collects the distance data of the nearest smoke in real time through the external visual system, and predicts the smoke diffusion speed and direction data in real time through the AI ​​model. The coordinate data of the drone is collected in real time through the GPS positioning system, and the distance data, smoke diffusion speed and direction data and coordinate data are shared in real time with all drones in the communication area through the communication system;

[0009] S3: The swarm intelligence algorithm is based on the ant algorithm and particle swarm optimization algorithm. It realizes the collaborative operation of drones through sharing information and collaborative decision-making. The main control unit of each drone adjusts the route in real time through the swarm intelligence algorithm based on the shared data, combined with its own real-time position, power and mission requirements. The swarm intelligence algorithm optimizes the flight paths between drones to prevent drones from entering areas with high concentrations of smoke and ensure the safe flight of drones;

[0010] S4: The ground command center includes a high-performance computing platform, a data processing module, and a communication module. The high-performance computing platform is used to process all the data fed back by the drone in real time and generate the optimal route planning plan through the optimization algorithm. The data processing module is responsible for data storage and analysis, and the communication module is responsible for real-time communication with the drone. The ground command center generates the optimal route planning plan through the optimization algorithm, combined with the real-time status of the drone and the dynamic changes of the wildfire, and sends the command to the drone through the communication module;

[0011] S5: Any drone that receives a command from the ground command center will share the command with all drones within the communication distance through the communication system. All drones adjust their routes according to the latest command to ensure the coordinated operation of the entire drone cluster. The communication system supports multi-point to multi-point communication mode to ensure the rapid transmission and sharing of commands.

[0012] The present invention integrates a high-precision GPS positioning system, an external visual system, and a communication system to achieve precise positioning of drones and real-time data sharing, thereby improving the response speed and accuracy of wildfire monitoring. The application of swarm intelligence algorithms enables drones to work together, optimize flight paths, avoid areas with high concentrations of smoke, and ensure flight safety.

[0013] Preferably, in S2, the AI ​​model adopts a convolutional neural network structure, specifically including an input layer, multiple convolutional layers, a pooling layer, a fully connected layer, and an output layer. The convolutional layer is used to extract the features of the smoke image, the pooling layer is used to reduce the dimension and the amount of calculation, the fully connected layer is used for classification and regression, and the output layer outputs the diffusion speed and direction of the smoke. The model is trained by a back-propagation algorithm, and the training data includes historical smoke images and their corresponding diffusion speeds and directions;

[0014] The AI ​​model of the convolutional neural network structure used in the present invention can efficiently extract smoke image features and predict the speed and direction of smoke diffusion, thereby improving the accuracy and prediction ability of smoke monitoring. The model is trained through the back propagation algorithm, so that the model can continuously learn and adapt to new data, thus improving the intelligence level of the monitoring system.

[0015] Preferably, in S2, the external vision system adopts a multimodal fusion system, and the external vision system specifically includes an optical sensor, an infrared sensor and a particle sensor. The optical sensor is used to capture smoke images, the infrared sensor is used to detect the position of the heat source in the smoke, and the particle sensor is used to detect the concentration of particulate matter in the smoke. The main control unit uses a fusion algorithm to comprehensively process the data of the multimodal sensors to improve the accuracy and reliability of the smoke distance and smoke diffusion speed and direction data.

[0016] The external vision system of the multimodal fusion system of the present invention can improve the accuracy and reliability of smoke monitoring through the integrated data of optical, infrared and particle sensors, so that the UAV can more comprehensively understand and respond to changes in the smoke environment.

[0017] Preferably, in S3, the ant algorithm adopts an adaptive step size strategy, specifically including:

[0018] Initial step size setting, step size adjustment factor and dynamic step size adjustment algorithm. The initial step size is set according to the initial position and mission requirements of the UAV. The step size adjustment factor is dynamically adjusted according to the real-time status of the UAV and the smoke distance data. The dynamic step size adjustment algorithm optimizes the step size in real time according to the collaborative information between UAVs to ensure efficient flight of the UAV in a smoke environment.

[0019] The present invention uses the adaptive step-size strategy of the ant algorithm to enable the UAV to dynamically adjust the flight step-size according to the real-time status and smoke distance data, thereby improving the flight efficiency and safety of the UAV in a complex smoke environment.

[0020] Preferably, the calculation formula of the step size adjustment factor is:

[0021]

[0022] Among them, α i is the step size adjustment factor of the i-th UAV;

[0023] α0 is the initial step length;

[0024] β is the adjustment coefficient;

[0025] C i is the smoke distance of the i-th UAV;

[0026] The update formula of the dynamic step size adjustment algorithm is:

[0027] α i =α i +γ·(P i -P i-1 );

[0028] Among them, γ is the learning rate;

[0029] P i is the current position of the i-th drone;

[0030] P i-1 is the previous position of the i-th UAV.

[0031] The present invention uses a specific step length adjustment factor calculation formula and a dynamic step length adjustment algorithm, so that the unmanned aerial vehicle can more accurately adjust the flight step length, optimize the flight route, and improve the flight efficiency and adaptability.

[0032] Preferably, in S3, the particle swarm optimization algorithm adopts a multi-objective optimization strategy, specifically including objective function design, weight factor setting and multi-objective optimization algorithm, the objective function design includes minimizing the probability of the UAV entering a high-concentration smoke area, minimizing the probability of collision between UAVs and maximizing the mission completion of the UAV, the weight factor is dynamically adjusted according to the real-time status of the UAV and the mission requirements, and the multi-objective optimization algorithm realizes the comprehensive optimization of multiple objectives through the particle swarm optimization algorithm to ensure the efficient flight of the UAV in complex terrain and smoke environment;

[0033] The objective function is:

[0034] f(x)=w1·P smoke +w2·P collision -w3·P task ;

[0035] Among them, P smoke is the probability of the drone entering a high-concentration smoke area;

[0036] P collision is the collision probability between drones;

[0037] P task Mission completion degree of the drone;

[0038] w1, w2, and w3 are the corresponding weight factors respectively.

[0039] The multi-objective optimization strategy of the particle swarm optimization algorithm of the present invention can maximize the task completion while ensuring the safety of the UAV, thereby improving the operating efficiency of the UAV in complex terrain and smog environments.

[0040] Preferably, in S4, the optimization algorithm adopts a genetic algorithm, specifically including a coding strategy, a selection strategy, a crossover strategy, a mutation strategy and a fitness function. The coding strategy adopts binary coding, the selection strategy adopts roulette selection, the crossover strategy adopts single-point crossover, the mutation strategy adopts bit flip mutation, and the fitness function is calculated according to the real-time status of the UAV and the dynamic changes of the wildfire. The optimal route planning scheme is generated through the genetic algorithm to ensure the efficient operation of the UAV in a complex environment.

[0041] The application of the genetic algorithm of the present invention makes the UAV route planning more flexible and efficient. Through the calculation of the fitness function, the optimal route planning scheme can be generated, thereby improving the operating efficiency of the UAV in complex environments.

[0042] Preferably, in S5, the communication system adopts multi-hop communication technology, specifically including a relay node selection algorithm, a routing algorithm and a data transmission protocol. The relay node selection algorithm selects the optimal relay node according to the real-time position and communication status of the drone, the routing algorithm generates the optimal communication path according to the cooperation information and communication status between the drones, and the data transmission protocol supports a multi-point to multi-point communication mode to ensure the rapid transmission and sharing of instructions.

[0043] The application of the multi-hop communication technology of the present invention improves the reliability and anti-interference capability of the communication system, ensures the rapid transmission and sharing of instructions, and improves the collaborative operation capability of the drone cluster.

[0044] Preferably, in S5, the communication system adopts coding redundancy technology, specifically including a data coding module, a redundant coding module and a decoding module. The data coding module encodes the instruction data, the redundant coding module inserts redundant codes into the encoded data, and the decoding module decodes and verifies the received data. Through coding redundancy technology, the reliability and anti-interference ability of instruction transmission are improved, ensuring the efficient operation of the drone in a complex environment.

[0045] The coding redundancy technology of the present invention improves the reliability of command transmission, ensures the efficient operation of the UAV in complex environments, and reduces the flight risks caused by communication errors.

[0046] Preferably, in S1, the GPS positioning system adopts a multi-constellation GNSS method, specifically including a GPS module, a GLONASS module, a Galileo module and a Beidou module. The GPS module is used to receive GPS satellite signals, the GLONASS module is used to receive GLONASS satellite signals, the Galileo module is used to receive Galileo satellite signals, and the Beidou module is used to receive Beidou satellite signals. Through the multi-constellation GNSS technology, the accuracy and reliability of the GPS positioning system are improved to ensure the efficient operation of the drone in a complex environment.

[0047] The multi-constellation GNSS technology of the present invention improves the accuracy and reliability of the GPS positioning system, enables the UAV to be positioned more accurately in complex environments, and improves the accuracy and safety of operations.

[0048] Beneficial effects:

[0049] The method realizes precise positioning, real-time data collection and efficient communication of the UAV by integrating a high-precision GPS positioning system, a multi-modal external vision system and a multi-hop communication system.

[0050] The AI ​​model using convolutional neural network structure and multimodal fusion system improves the accuracy and predictive ability of smoke monitoring.

[0051] The application of swarm intelligence algorithms and genetic algorithms enables drones to adjust their routes in real time, optimize flight paths, avoid areas with high concentrations of smoke, and ensure flight safety.

[0052] The application of coding redundancy technology and multi-constellation GNSS technology improves the reliability of communication and positioning accuracy, enabling drones to perform wildfire monitoring tasks efficiently and safely in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0054] Figure 1 is a flow chart of the steps of the present invention;

[0055] Figure 2 is a schematic block diagram of an external visual system of the present invention;

[0056] Figure 3 It is a flowchart of the adaptive step size strategy. DETAILED DESCRIPTION

[0057] In order to enable those skilled in the art to better understand the solution of the present application, the following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0058] Example

[0059] The method for planning the route of a wildfire monitoring UAV based on real-time smoke dynamics includes the following steps:

[0060] S1: UAV preparation. When encountering a wildfire, at least two UAVs will be launched at the same time. Any UAV is equipped with a GPS positioning system, an external vision system, and a communication system. The GPS positioning system includes a high-precision GPS antenna and a receiving module to obtain the precise position of the UAV in real time. The external vision system is used to collect smoke images in real time. The communication system includes a variety of communication modules to support long-distance wireless communication and short-distance high-speed data transmission.

[0061] S2: Data collection and analysis: any drone collects the distance data of the nearest smoke in real time through the external visual system, and predicts the smoke diffusion speed and direction data in real time through the AI ​​model. The coordinate data of the drone is collected in real time through the GPS positioning system, and the distance data, smoke diffusion speed and direction data and coordinate data are shared in real time with all drones in the communication area through the communication system;

[0062] S3: The swarm intelligence algorithm is based on the ant algorithm and particle swarm optimization algorithm. It realizes the collaborative operation of drones through sharing information and collaborative decision-making. The main control unit of each drone adjusts the route in real time through the swarm intelligence algorithm based on the shared data, combined with its own real-time position, power and mission requirements. The swarm intelligence algorithm optimizes the flight paths between drones to prevent drones from entering areas with high concentrations of smoke and ensure the safe flight of drones;

[0063] S4: The ground command center includes a high-performance computing platform, a data processing module, and a communication module. The high-performance computing platform is used to process all the data fed back by the drone in real time and generate the optimal route planning plan through the optimization algorithm. The data processing module is responsible for data storage and analysis, and the communication module is responsible for real-time communication with the drone. The ground command center generates the optimal route planning plan through the optimization algorithm, combined with the real-time status of the drone and the dynamic changes of the wildfire, and sends the command to the drone through the communication module;

[0064] S5: Any drone that receives a command from the ground command center will share the command with all drones within the communication distance through the communication system. All drones adjust their routes according to the latest command to ensure the coordinated operation of the entire drone cluster. The communication system supports multi-point to multi-point communication mode to ensure the rapid transmission and sharing of commands.

[0065] The present invention integrates a high-precision GPS positioning system, an external visual system, and a communication system to achieve precise positioning of drones and real-time data sharing, thereby improving the response speed and accuracy of wildfire monitoring. The application of swarm intelligence algorithms enables drones to work together, optimize flight paths, avoid areas with high concentrations of smoke, and ensure flight safety.

[0066] Preferably, in S2, the AI ​​model adopts a convolutional neural network structure, specifically including an input layer, multiple convolutional layers, a pooling layer, a fully connected layer, and an output layer. The convolutional layer is used to extract the features of the smoke image, the pooling layer is used to reduce the dimension and the amount of calculation, the fully connected layer is used for classification and regression, and the output layer outputs the diffusion speed and direction of the smoke. The model is trained by a back-propagation algorithm, and the training data includes historical smoke images and their corresponding diffusion speeds and directions;

[0067] The AI ​​model of the convolutional neural network structure used in the present invention can efficiently extract smoke image features and predict the speed and direction of smoke diffusion, thereby improving the accuracy and prediction ability of smoke monitoring. The model is trained through the back-propagation algorithm, so that the model can continuously learn and adapt to new data, thus improving the intelligence level of the monitoring system.

[0068] In this embodiment, in S2, the external vision system adopts a multimodal fusion system, and the external vision system specifically includes an optical sensor, an infrared sensor and a particle sensor. The optical sensor is used to capture smoke images, the infrared sensor is used to detect the position of the heat source in the smoke, and the particle sensor is used to detect the concentration of particulate matter in the smoke. The main control unit uses a fusion algorithm to comprehensively process the data of the multimodal sensors to improve the accuracy and reliability of the smoke distance and smoke diffusion speed and direction data.

[0069] The external vision system of the multimodal fusion system of the present invention can improve the accuracy and reliability of smoke monitoring through the integrated data of optical, infrared and particle sensors, so that the UAV can more comprehensively understand and respond to changes in the smoke environment.

[0070] In this embodiment, in S3, the ant algorithm adopts an adaptive step size strategy, which specifically includes:

[0071] Initial step size setting, step size adjustment factor and dynamic step size adjustment algorithm. The initial step size is set according to the initial position and mission requirements of the UAV. The step size adjustment factor is dynamically adjusted according to the real-time status of the UAV and the smoke distance data. The dynamic step size adjustment algorithm optimizes the step size in real time according to the collaborative information between UAVs to ensure efficient flight of the UAV in a smoke environment.

[0072] The present invention uses the adaptive step-size strategy of the ant algorithm to enable the UAV to dynamically adjust the flight step-size according to the real-time status and smoke distance data, thereby improving the flight efficiency and safety of the UAV in a complex smoke environment.

[0073] In this embodiment, the calculation formula of the step size adjustment factor is:

[0074]

[0075] Among them, αi is the step size adjustment factor of the i-th UAV;

[0076] α0 is the initial step length;

[0077] β is the adjustment coefficient;

[0078] C i is the smoke distance of the i-th UAV;

[0079] The update formula of the dynamic step size adjustment algorithm is:

[0080] α i =α i +γ·(P i -P i-1 );

[0081] Among them, γ is the learning rate;

[0082] p i is the current position of the i-th drone;

[0083] P i-1 is the previous position of the i-th UAV.

[0084] The present invention uses a specific step length adjustment factor calculation formula and a dynamic step length adjustment algorithm, so that the unmanned aerial vehicle can more accurately adjust the flight step length, optimize the flight route, and improve the flight efficiency and adaptability.

[0085] In this embodiment, in S3, the particle swarm optimization algorithm adopts a multi-objective optimization strategy, specifically including objective function design, weight factor setting and multi-objective optimization algorithm. The objective function design includes minimizing the probability of the UAV entering the high-concentration smoke area, minimizing the collision probability between UAVs and maximizing the task completion of the UAV. The weight factor is dynamically adjusted according to the real-time status of the UAV and the task requirements. The multi-objective optimization algorithm realizes the comprehensive optimization of multiple objectives through the particle swarm optimization algorithm to ensure the efficient flight of the UAV in complex terrain and smoke environment.

[0086] The objective function is:

[0087] f(x)=w1·P smoke +w2·P collision -w3·P task ;

[0088] Among them, P smoke is the probability of the drone entering a high-concentration smoke area;

[0089] P collision is the collision probability between drones;

[0090] P task Mission completion degree of the drone;

[0091] w1, w2, and w3 are the corresponding weight factors respectively.

[0092] The multi-objective optimization strategy of the particle swarm optimization algorithm of the present invention can maximize the task completion while ensuring the safety of the UAV, thereby improving the operating efficiency of the UAV in complex terrain and smog environments.

[0093] In this embodiment, in S4, the optimization algorithm adopts a genetic algorithm, specifically including a coding strategy, a selection strategy, a crossover strategy, a mutation strategy and a fitness function. The coding strategy adopts binary coding, the selection strategy adopts roulette selection, the crossover strategy adopts single-point crossover, and the mutation strategy adopts bit flip mutation. The fitness function is calculated according to the real-time status of the UAV and the dynamic changes of the wildfire. The optimal route planning scheme is generated through the genetic algorithm to ensure the efficient operation of the UAV in a complex environment.

[0094] The application of the genetic algorithm of the present invention makes the UAV route planning more flexible and efficient. Through the calculation of the fitness function, the optimal route planning scheme can be generated, thereby improving the operating efficiency of the UAV in complex environments.

[0095] In this embodiment, in S5, the communication system adopts multi-hop communication technology, specifically including relay node selection algorithm, routing algorithm and data transmission protocol. The relay node selection algorithm selects the optimal relay node according to the real-time position and communication status of the drone, the routing algorithm generates the optimal communication path according to the cooperation information and communication status between drones, and the data transmission protocol supports multi-point to multi-point communication mode to ensure the rapid transmission and sharing of instructions.

[0096] The application of the multi-hop communication technology of the present invention improves the reliability and anti-interference capability of the communication system, ensures the rapid transmission and sharing of instructions, and improves the collaborative operation capability of the drone cluster.

[0097] In this embodiment, in S5, the communication system adopts coding redundancy technology, which specifically includes a data coding module, a redundant coding module and a decoding module. The data coding module encodes the instruction data, the redundant coding module inserts redundant codes into the encoded data, and the decoding module decodes and verifies the received data. Through coding redundancy technology, the reliability and anti-interference ability of instruction transmission are improved, ensuring the efficient operation of the drone in complex environments.

[0098] The coding redundancy technology of the present invention improves the reliability of command transmission, ensures the efficient operation of the UAV in complex environments, and reduces the flight risks caused by communication errors.

[0099] In this embodiment, in S1, the GPS positioning system adopts a multi-constellation GNSS method, specifically including a GPS module, a GLONASS module, a Galileo module and a Beidou module. The GPS module is used to receive GPS satellite signals, the GLONASS module is used to receive GLONASS satellite signals, the Galileo module is used to receive Galileo satellite signals, and the Beidou module is used to receive Beidou satellite signals. Through the multi-constellation GNSS technology, the accuracy and reliability of the GPS positioning system are improved to ensure the efficient operation of the drone in a complex environment.

[0100] The multi-constellation GNSS technology of the present invention improves the accuracy and reliability of the GPS positioning system, enables the UAV to be positioned more accurately in complex environments, and improves the accuracy and safety of operations.

[0101] The optional implementation modes of the embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation modes. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical scheme of the embodiments of the present invention, and these simple modifications all belong to the protection scope of the embodiments of the present invention.

[0102] The operating principle of the present invention is as follows:

[0103] When a wildfire occurs, the ground command center starts the drone takeoff procedure and takes off at least two drones at the same time. Before each drone takes off, the main control unit activates the GPS positioning system, external vision system, communication system and smoke concentration sensor. After the drone takes off, the external vision system collects smoke images in real time, the smoke concentration sensor measures the smoke concentration in real time, and the GPS positioning system collects the coordinate data of the drone in real time. All data are transmitted to other drones and the ground command center in real time through the communication system. The high-performance computing platform of the ground command center processes the data fed back by all drones in real time and generates the optimal route planning plan through the optimization algorithm. The ground command center sends the command to the drone through the communication module, and any drone that receives the command shares the command with other drones through the communication system. All drones adjust their routes in real time according to the latest instructions through the swarm intelligence algorithm to avoid entering high-concentration smoke areas and ensure the safe flight of the drones. The main control unit of each drone adjusts the route in real time through the swarm intelligence algorithm based on the shared data, combined with its own real-time position, power and mission requirements. The swarm intelligence algorithm optimizes the flight paths between drones to ensure the collaborative operation between drones and improve the flight efficiency and safety of drones in complex terrain and smoke environments. In data collection and analysis, the camera in the external vision system collects smoke images in real time, and the image processing module extracts the distance data of the smoke through the convolutional neural network. The smoke concentration sensor measures the concentration of smoke in real time through multimodal fusion technology and transmits the data to the main control unit of the drone. The AI ​​model is trained with historical data to predict the diffusion speed and direction of smoke in real time. The GPS positioning system collects the coordinate data of the drone in real time and shares all the data with other drones through the communication system. The main control unit stores and processes the data collected by the sensor through the central processor, data storage and signal transceiver, and transmits the data to other drones through the communication system. In the swarm intelligence algorithm, the swarm intelligence algorithm is based on the ant algorithm and the particle swarm optimization algorithm, and realizes the collaborative operation of drones through sharing information and collaborative decision-making. The main control unit of each drone adjusts the route in real time through the swarm intelligence algorithm based on the shared data, combined with its own real-time position, power and mission requirements. The adaptive step size strategy and multi-objective optimization strategy ensure the efficient flight of the drone in the smoke environment. In the ground command center, the high-performance computing platform processes all the data fed back by the drone in real time through genetic algorithms and deep learning models to generate the optimal route planning plan. The data processing module extracts data features through convolutional neural networks, and the optimization algorithm generates the optimal route planning scheme through genetic algorithms. The optimization algorithm combines the real-time status of the drone and the dynamic changes of the wildfire to generate the optimal route planning scheme, and sends the instructions to the drone through the communication module. In command sharing, the communication system uses multi-hop communication technology and coding redundancy technology to ensure the rapid transmission and sharing of instructions.The relay node selection algorithm and routing algorithm generate the optimal communication path, and the data transmission protocol supports multi-point to multi-point communication mode. The coding redundancy technology improves the reliability and anti-interference ability of instruction transmission through data coding module, redundant coding module and decoding module.

[0104] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

[0105] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including several instructions for making a single-chip microcomputer, a chip or a processor (processor) execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0106] In addition, various implementation modes of the embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed by the embodiments of the present invention.

Claims

1. A method for planning the route of a wildfire monitoring drone based on real-time smoke dynamics, characterized in that: The following steps are involved: S1: UAV preparation. When encountering a wildfire, at least two UAVs are launched at the same time. Any UAV is equipped with a GPS positioning system, an external vision system, and a communication system. The GPS positioning system includes a high-precision GPS antenna and a receiving module to obtain the precise position of the UAV in real time. The external vision system is used to collect smoke images in real time. The communication system includes a variety of communication modules to support long-distance wireless communication and short-distance high-speed data transmission. S2: Data collection and analysis: any drone collects the distance data of the nearest smoke in real time through the external visual system, and predicts the smoke diffusion speed and direction data in real time through the AI ​​model. The coordinate data of the drone is collected in real time through the GPS positioning system, and the distance data, smoke diffusion speed and direction data and coordinate data are shared in real time with all drones in the communication area through the communication system; S3: The swarm intelligence algorithm is based on the ant algorithm and particle swarm optimization algorithm. It realizes the collaborative operation of drones through sharing information and collaborative decision-making. The main control unit of each drone adjusts the route in real time through the swarm intelligence algorithm based on the shared data, combined with its own real-time position, power and mission requirements. The swarm intelligence algorithm optimizes the flight paths between drones to prevent drones from entering high-concentration smoke areas and ensure the safe flight of drones. S4: The ground command center includes a high-performance computing platform, a data processing module, and a communication module. The high-performance computing platform is used to process all the data fed back by the drone in real time and generate the optimal route planning plan through the optimization algorithm. The data processing module is responsible for data storage and analysis, and the communication module is responsible for real-time communication with the drone. The ground command center generates the optimal route planning plan through the optimization algorithm, combined with the real-time status of the drone and the dynamic changes of the wildfire, and sends the command to the drone through the communication module; S5: Any drone that receives a command from the ground command center will share the command with all drones within the communication distance through the communication system. All drones will adjust their routes according to the latest command to ensure the collaborative operation of the entire drone cluster. The communication system supports multi-point to multi-point communication mode to ensure the rapid transmission and sharing of commands.

2. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 1 is characterized in that: In S2, the AI ​​model adopts a convolutional neural network structure, specifically including an input layer, multiple convolutional layers, a pooling layer, a fully connected layer and an output layer. The convolutional layer is used to extract the features of the smoke image, the pooling layer is used to reduce the dimension and reduce the amount of calculation, the fully connected layer is used for classification and regression, and the output layer outputs the diffusion speed and direction of the smoke. The model is trained through the back propagation algorithm, and the training data includes historical smoke images and their corresponding diffusion speeds and directions.

3. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 1 is characterized in that: In S2, the external vision system adopts a multimodal fusion system. The external vision system specifically includes optical sensors, infrared sensors and particle sensors. The optical sensor is used to capture smoke images, the infrared sensor is used to detect the position of heat sources in the smoke, and the particle sensor is used to detect the concentration of particulate matter in the smoke. The main control unit uses a fusion algorithm to comprehensively process the data of multimodal sensors to improve the accuracy and reliability of smoke distance, smoke diffusion speed and direction data.

4. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 1 is characterized in that: In S3, the ant algorithm adopts an adaptive step size strategy, which specifically includes: Initial step size setting, step size adjustment factor and dynamic step size adjustment algorithm. The initial step size is set according to the initial position and mission requirements of the UAV. The step size adjustment factor is dynamically adjusted according to the real-time status of the UAV and the smoke distance data. The dynamic step size adjustment algorithm optimizes the step size in real time according to the collaborative information between UAVs to ensure efficient flight of the UAV in a smoke environment.

5. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 4 is characterized in that: The calculation formula of the step size adjustment factor is: Among them, α i is the step size adjustment factor of the i-th UAV; α0 is the initial step length; β is the adjustment coefficient; C i is the smoke distance of the i-th UAV; The update formula of the dynamic step size adjustment algorithm is: a i =a i +γ·(P i -P i-1 ); Among them, γ is the learning rate; p i is the current position of the i-th drone; P i-1 is the previous position of the i-th UAV.

6. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 4 is characterized in that: In S3, the particle swarm optimization algorithm adopts a multi-objective optimization strategy, specifically including objective function design, weight factor setting and multi-objective optimization algorithm. The objective function design includes minimizing the probability of the UAV entering the high-concentration smoke area, minimizing the collision probability between UAVs and maximizing the task completion of the UAV. The weight factor is dynamically adjusted according to the real-time status of the UAV and the task requirements. The multi-objective optimization algorithm realizes the comprehensive optimization of multiple objectives through the particle swarm optimization algorithm to ensure the efficient flight of the UAV in complex terrain and smoke environment. The objective function is: f(x)=w1·P smoke +w2·P collision -w3·P task ; Among them, P smoke is the probability of the drone entering a high-concentration smoke area; P collision is the collision probability between drones; P task Mission completion degree of the drone; w1, w2, and w3 are the corresponding weight factors respectively.

7. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 1 is characterized in that: In S4, the optimization algorithm adopts a genetic algorithm, which specifically includes a coding strategy, a selection strategy, a crossover strategy, a mutation strategy and a fitness function. The coding strategy adopts binary coding, the selection strategy adopts roulette selection, the crossover strategy adopts single-point crossover, and the mutation strategy adopts bit flip mutation. The fitness function is calculated based on the real-time status of the UAV and the dynamic changes of the wildfire. The optimal route planning scheme is generated through the genetic algorithm to ensure the efficient operation of the UAV in a complex environment.

8. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 1 is characterized in that: In S5, the communication system adopts multi-hop communication technology, which specifically includes a relay node selection algorithm, a routing algorithm and a data transmission protocol. The relay node selection algorithm selects the optimal relay node according to the real-time position and communication status of the drone. The routing algorithm generates the optimal communication path according to the collaborative information and communication status between the drones. The data transmission protocol supports a multi-point to multi-point communication mode to ensure the rapid transmission and sharing of instructions.

9. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 8 is characterized in that: In S5, the communication system adopts coding redundancy technology, which specifically includes a data coding module, a redundant coding module and a decoding module. The data coding module encodes the instruction data, the redundant coding module inserts redundant codes into the encoded data, and the decoding module decodes and verifies the received data. Through the coding redundancy technology, the reliability and anti-interference capability of instruction transmission are improved, ensuring the efficient operation of the UAV in complex environments.

10. The method for wildfire monitoring UAV route planning based on real-time smoke dynamics according to claim 1 is characterized in that: In S1, the GPS positioning system adopts a multi-constellation GNSS method, specifically including a GPS module, a GLONASS module, a Galileo module and a Beidou module. The GPS module is used to receive GPS satellite signals, the GLONASS module is used to receive GLONASS satellite signals, the Galileo module is used to receive Galileo satellite signals, and the Beidou module is used to receive Beidou satellite signals. Through the multi-constellation GNSS technology, the accuracy and reliability of the GPS positioning system are improved to ensure the efficient operation of the drone in complex environments.

Images